5242
S. R. Bavikar et al. / Tetrahedron Letters 55 (2014) 5241–5243
Table 1
Table 2
Selective reduction of compounds 1
Selective reduction of compounds 1j–1m
Entry Reactant
Product
Yielda (%)
Entry
Reactant
O
Product
Yielda (%)
78
O
O
O
O
O
O
O
1
2
O
O
HO
O
2 e
HO
1 j
1
2
3
45
O
O
HO
OHC
O
O
67
43
O
2 e
2 a
1 a
O
1 k
O
OH
O
O
O
O
O
NO2
NO2
O
O
3
4
HO
O
HO
O
HO
1l
2 f
89
O
OH
O
O
75
O
O
1 b
2 b
1 m
2 g
O
O
O
O
O
a
Isolated yield; variable amount of starting material was recovered.
O
O
HO
Based on the reported information,2 the substituted cyclopente-
O
55
none alcohol 1a was subjected to Luche reduction with an aim to
obtain the corresponding diol. There was no reaction with one
equivalent of sodium borohydride and cerium chloride so the reac-
tion was attempted with varying amounts of sodium borohydride
and cerium chloride. To our surprise, when the reaction was car-
ried out with 2 equiv of cerium chloride and 3 equiv of sodium
borohydride, the substituted cyclopentenone 2a was obtained in
45% yield (Scheme 2) and rest of the starting material was
recovered.
O
O
OH
2 c
O
1 c
O
O
O
O
O
O
O
O
Si
O
4
5
50
85
Thus, the ketone carbonyl in 1a survived while the hydroxyl
functionality was reductively removed and aldehyde was reduced
to alcohol to afford the substituted cyclopentenone 2a.
O
O
2 b
1 d
This unprecedented interesting result prompted us to check the
generality of the reaction and a number of cyclopentenone alcohols
were subjected to Luche reduction (Scheme 3).
O
O
O
O
O
O
HO
It was observed that in all the above cases,14,15 deoxygenation
of cyclopentenone alcohols 1 took place to give cyclopentenones
2 in 2–3 h, instead of the reduction of ketone functionality, in good
yields (Table 1).
O
O
O
O
2 d
1 e
O
O
O
O
O
6
7
55
51
The reaction was further studied for its scope and limitations.
When 4-hydroxy-2-phenylcyclopent-2-en-1-one (1j) was sub-
jected to similar reaction (Table 2, entry 1), 2-phenylcyclopent-2-
en-1-one (2e) was obtained in 78% yield while the reaction of 4-
acetoxy-2-phenylcyclopent-2-en-1-one (1k) afforded 2-phenylcy-
clopent-2-en-1-one (2e) in 67% yield (Table 2, entry 2). The reac-
tion of 4-hydroxy-2-(4-nitrophenyl)cyclopent-2-en-1-one (1l)
was slow and incomplete and resulted in the formation of 1,4-
dihydroxy-2-(4-nitrophenyl)cyclopent-2-ene (2f) in 43% yield
after 4 h. The corresponding 4-dehydroxylated product 2-(4-nitro-
phenyl) cyclopent-2-en-1-one was not obtained. Similarly, 4-acet-
Si
O
O
O
2 d
1 f
O
O
O
O
O
O
O
O
O
O
O
O
2 d
2 d
1 g
O
O
O
O
O
8
65
O
O
O
oxycyclopent-2-en-1-one
(1m)
provided
1-acetoxy-4-
1 h
hydroxycyclopent-2-ene (2g) in 75% yield and cyclopent-2-en-1-
one was not obtained.
O
O
O
Thus, it was observed that the 4-(un)protectedhydroxycyclo-
pent-2-en-1-ones bearing phenyl ring with electron donating sub-
stituent or unsubstituted phenyl ring, at 2-position of
cyclopentenone undergo deoxygenation at 4-position while 4-
(un)protectedhydroxycyclopent-2-en-1-ones bearing phenyl ring,
with electron withdrawing substituent (e.g., nitro group), at 2-
position of cyclopentenone undergo usual reduction of ketone
functionality. In case of substrates with same substituents on aro-
O
O
Si
9
49
O
O
O
2 d
1i
Isolated yield; variable amount of starting material was recovered.
a